Intelligent Voice Question Answering System for Agricultural Production Based on Deep Learning

Agricultural production is the main source of food, providing humans with various grains, vegetables, fruits, meat, and other nutrients. Ensuring sufficient food supply is the basic condition for maintaining human health and survival. Agriculture is one of the important pillars of the national economy. It provides the country with abundant agricultural products and raw materials, creates employment opportunities, promotes the development of rural economy, and promotes the increase of farmers’ income and the development of rural areas. Therefore, agricultural production is important in many aspects and has a profound impact on social stability and human survival and development. But, the current agricultural production has important flaws in terms of intelligence. In recent years, artificial intelligence technology represented by machine learning and deep learning has advanced by leaps and bounds. While these technologies have changed our lives, they have also caused an explosive growth of various information. How can we quickly and accurately find what we need from a large amount of information? The answer to the question has become a widespread concern of relevant researchers. At present, the question answering technology basically adopts the retrieval-reading comprehension interactive framework, but there are still many defects in the agricultural question answering system based on this architecture, which makes that the answers provided by the model may not be concise and clear, and this paper will improve the problems existing in the current agricultural question answering technology. This paper proposes a question answering algorithm based on multi-task learning. This paper first enables the model to capture multiple. At the same time, it can better learn some features that are difficult to capture in a single task and improve the ability of the model through attention, so that the answers generated by it are more concise and clear.